Quickly discharge power-supply capacitors
Edited by Bill Travis and Anne Watson Swager
Stephen Woodward, University of North Carolina, Chapel Hill, NC -- EDN, July 5, 2001
A perennial challenge in power-supply design is the safe and speedy discharge, or "dump," at turn-off of the large amount of energy stored in the postrectification filter capacitors. This energy, CV2/2, can usually reach tens of joules. If you let the capacitors self-discharge, dangerous voltages can persist on unloaded electrolytic filter capacitors for hours or even days. These charged capacitors can pose a significant hazard to service personnel or even to the equipment itself. The standard and obvious solution to this problem is the traditional "bleeder" resistor, RB (Figure 1). The trouble with the RB fix is that power continuously and wastefully "bleeds" through RB, not only when it's desirable during a capacitor dump, but also constantly when the power supply is on. The resulting energy hemorrhage is sometimes far from negligible.
Figure 1 offers an illustration of the problem, taken from the power supply of a pulse generator. The CV2/2 energy stored at the nominal 150V operating voltage is 1502×4400 µF/2, or approximately 50J. Suppose you choose the RB fix for this supply and opt to achieve 90% discharge of the 4400-µF capacitor within 10 sec after turning off the supply. You then have to select RB to provide a constant RC time no longer than 10/ln(10), or 4.3 sec. RB, therefore, equals 4.3 sec/4400 µF, or approximately 1 kV. The resulting continuous power dissipated in RB is 1502/1 kV, or approximately 23W. This figure represents an undesirable power-dissipation penalty in a low-duty-cycle pulse-generator application. This waste dominates all energy consumption and heat production in what is otherwise a low-average-power circuit. This scenario is an unavoidable drawback of bleeder resistors. Whenever you apply the 10%-in-10-sec safety criterion, the downside is the inevitable dissipation of almost half the CV2/2 energy during each second the circuit is under power.
Figure 2 shows a much more selective and thrifty fix for the energy-dump problem. The otherwise-unused off-throw contacts of the DPDT on/off power switch create a filter-capacitor-discharge path that exists only when you need it: when the supply is turned off. When the switch moves to the off position, it establishes a discharge path through resistors R1 and R2 and the power transformer's primary winding. The result is an almost arbitrarily rapid dump of the stored energy, while the circuit suffers zero power-on energy waste. Use the following four criteria to optimally select R1, R2, and S1:
-
The peak discharge current, V/(R1+R2), should not exceed S1's contact rating.
-
The pulse-handling capability of R1 and R2 should be adequate to handle the CV2/2 thermal impulse. A 3W rating for R1 and R2 is adequate for this 50J example.
-
The discharge time constant, (R1+R2)C, should be short enough to ensure quick disposal of the stored energy.
-
S1 must have a break-before-make architecture that ensures breaking both connections to the ac mains before making either discharge connection, and vice versa. Otherwise, a hazardous ground-fault condition may occur at on/off transitions.
-
This idea is not good if you are using this to satisfy a safety regulation, because if the supply is unplugged with the power switch left on, the safety feature will not function. This would be common, for example for power supplies plugged into a switched power strip.
Brian Park - 2008-24-12 14:15:00 PST
